The theft of opioids such as fentanyl and oxycodone by health care employees is a massive problem gripping U.S. hospitals and serves as a major pipeline for the drugs fueling the national opioid use disorder epidemic. Seeking new solutions, some health care systems are looking to technology, particularly artificial intelligence, to better, and more quickly, identify employees stealing such medications, Quartz reports.

The sharing of injection equipment, whether for the use of opioids or other narcotics, is driving the recent rise in new hepatitis C virus (HCV) cases, especially among young people. Increasing rates of behavior as a result of the swelling opioid epidemic is also apparently responsible for a reversal in the 15-year decline in HIV transmissions among people who inject drugs.

Drug diversion—a term for the theft of prescription drugs—in health care settings may lead patients to suffer if the pain medication they were supposed to receive is surreptitiously swapped with another substance, such as saline or over-the-counter drugs.

According to one low-ball estimate in the Quartz article, the country sees perhaps 37,000 drug diversion cases in the United States annually. According to another estimate, as many as 10% of pharmacists, nurses and anesthesiologists commit drug diversion.

Several companies offering technological solutions to this crisis have sprung up recently. Kit Check, for example, uses a machine-learning algorithm that apparently can point to risky prescribing patterns among health care providers, while another, Invistics Corp., relies on cloud-based controlled substance tracking software.

Critics of these programs characterize them as imperfect and potentially toothless, should hospitals not take the proper action in response to warnings of drug diversion by their staff.

That said, a recent wave of drug diversion scandals in prestigious health care systems, including the University of Michigan Health System and Massachusetts General Hospital, has put other health care providers on higher alert. The two systems were fined by the Drug Enforcement Agency (DEA) for violating provisions of the Controlled Substances Act by failing to engage in sufficient efforts to prevent and respond to drug diversion.

Mass General, which is affiliated with Harvard Medical School, is under orders from the federal government to step up its drug-diversion-fighting game. The health care system is now using a non-AI-powered automated system to track medications and aims to get an AI program up and running by 2021.

To read the Quartz article, click here.